End of training

README.md

@@ -21,7 +21,7 @@ model-index:
     metrics:
     - name: Wer
       type: wer
-      value: 48.38265205695164
+      value: 48.704815730601666
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,9 +31,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Darija STT Mix dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6476
-- Wer Ortho: 53.1133
-- Wer: 48.3827
+- Loss: 0.6617
+- Wer Ortho: 52.9142
+- Wer: 48.7048
 
 ## Model description
 
@@ -61,16 +61,15 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 200
-- training_steps: 1500
+- training_steps: 1000
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss | Wer Ortho | Wer     |
 |:-------------:|:------:|:----:|:---------------:|:---------:|:-------:|
-| 0.7485        | 0.8403 | 500  | 0.7254          | 63.9656   | 58.6313 |
-| 0.4356        | 1.6807 | 1000 | 0.6557          | 51.7551   | 46.4340 |
-| 0.2405        | 2.5210 | 1500 | 0.6476          | 53.1133   | 48.3827 |
+| 0.6715        | 0.8203 | 500  | 0.7303          | 64.8593   | 60.5440 |
+| 0.4268        | 1.6407 | 1000 | 0.6617          | 52.9142   | 48.7048 |
 
 
 ### Framework versions